496 PROCEEDINGS OF THE AMERICAN ACADEMY. 



lative in the case of 25 country names, some being the same as those 

 which he has used without the preposition. When Ussing remarks 

 ' Even the genitive appears,' he must mean the ' locative ' for there 

 would be nothing surprising in the employment of a true genitive con- 

 struction. I do not believe that the true locative of any country name is 

 found in Vitruvius, since I think that all the forms which seem to be 

 such may be explained on other grounds, just as the apparent locatives 

 of country names in Pliny have been explained away. 34 Oidy six cases 

 call for considei-ation. Of these Asiae (190, 14) and Phrygiae (196, 14) 

 are chorographic genitives (see Schmalz, Lat. Gr., z 234 f.), such as are 

 found in Caesar, Livy, and Pliny. In 195, 15 Aethiopiae is now read 

 Aethiopia, but, if the manuscript reading is kept, we have a genitive de- 

 pending on lacus. In 198, 8 Boeotiae (a genitive) has been emended to 

 Boeotia on account of the following ablatives. In the example cited by 

 Ussing, Cretae et Africae (59, 3), one of two explanations may be given. 

 Although the name Cretae is generally treated like that of a country and 

 consequently appears in Cicero with in and the ablative, yet as an island 

 it is used in the locative by Varro (R. P. 1, 7, 6) and Virgil (A. 3, 162). 

 If Vitruvius used it thus, then the following Africae is an assimilation 

 for concinnity, like Sallust's Romae Numidiaeque {J. 33, 4). But both 

 Cretae and Africae may be genitives depending on regionibus, for the 

 whole sentence reads : nascuntur autem eae arbores maxime Cretae et 

 Africae et nonnullis Syriae regionibus. There remains only 200, 24 : 

 sunt autem etiam fontes uti vino mixti, quemadmodum est unus Paphlago- 

 niae, ex quo, etc. Here Papldagoniae is to be taken as a genitive. But 

 even if locatives of country names were actually found in Vitruvius, we 

 could parallel them from the classical period, since we have Peloponnesi 

 in Varro (R. R. 2, 6, 2), Chersonesi in Nepos (1, 2, 4), and Galliae in 

 Hiitius (B. G. 8, 1, 2). 35 As for names of towns in the ablative instead 

 of the locative, Nohl's treatment (Analecta, p. 10) is not exact, for he 

 does not distinguish between towns and islands. The names of towns 

 actually thus used by Vitruvius are Arretio, Chio (283, 3 where the word 

 murum shows that the town is meant), Halicarnasso, Lyncesto, Paraetonio, 

 Sunio, Ta?so, Teo, Teano, — that is, nine in all. 36 It is true that this 

 misuse becomes common in late Latin {Archiv, XIII, 315 f.), but still we 



s* Funaioli, Archiv, 13, 581 f. 



35 Here I think that Galliae must certainly be taken as a locative on account 

 otreb'ts gestis Alexandriae just below. Still, see Archiv, 13, 831. 



36 For the passages, see Nohl. On the other hand Vitruvius has the locative 

 of stems in -o- six times, and always in stems in -a-. 



